Polanyi's paradox
The principle that we always know more than we can tell, which historically grounded the limit of automating complex cognitive tasks.
Michael Polanyi formulated the paradox in his work on tacit knowledge, distilled in The Tacit Dimension (1966). It fits in one sentence: we always know more than we can tell. A large share of human competence, from the surgeon's dexterity to the judge's discernment, rests on embodied, contextual and hard-to-formalize knowledge that no explicit rule describes in full. In automation, this was long read as a structural barrier to replacing skilled workers with algorithms: a task that cannot be entirely broken into explicit instructions cannot be handed to a machine. David Autor and co-authors formalized the intuition in the ALM model (2003), which expected routine tasks to be automated and left complex cognitive tasks to humans. Large language models and agentic AI get around the paradox without solving the philosophical problem: they produce expert-quality reasoning by statistical means, with no explicit instructions. That empirical workaround is the break that current research on the economic impact of AI on cognitive labor markets sets out to theorize.
Hypothetical case: an experienced marine insurance underwriter is unable to fully explain why he declines a risk whose objective data look acceptable: he draws on an intuition shaped by twenty years of experience. This is precisely the kind of tacit judgment that Polanyi's paradox, formulated in 1966, identified as beyond automation and that foundation models have partially learned to emulate through probabilistic approaches.
Polanyi paradox, tacit knowledge
Follow the research
One letter a week. First, the week's underwriting note, when one is published: a dated, argued decision, signed by Paul Bernardski. Then what the observatory published on cyber risk, political risk and insurance. No account, no tracking.
No spam, one-click unsubscribe, no data ever sold.